The primary objective of this work is to optimize the lift-to-drag ratio of a waverider-based configuration by a new global optimization method — the adaptive surrogate model (ASM), which is based on latin hypercube computer experiment and kriging surrogate model. Additional design points will be added in experiment set of points during iterative process. Design spaces of each variable are reduced by an adaptive reduction radius, which is improved gradually by the adaptive inconsistency of optimum solutions during the optimization process. Also the search efficiency and the accuracy of the optimization are compared with another global optimization scheme. At last, this paper gives maximum L/D optimization with restrictions of actual volumetric efficiency and the total mass, which shows that the adaptive surrogate model is quite suitable for the design optimization of waveriders.